This project uses a comprehensive longitudinal dataset to examine the complex multi-level processes underlying black children's disproportionate underutilization of mental health services and overrepresentation in special education relative to white children. Four specific aims are addressed: 1) To describe and analyze the dynamics of disparities in special education placement and mental health services use among black and white children; 2) To evaluate the degree to which child and family characteristics explain black/white disparity in special education entry and exit and in placement restrictiveness, and explain mental health services use; 3) To examine the influence of school and community characteristics on racial disparities in special education and mental health services use; and 4) To evaluate whether the process determining special education placement and use of mental health services varies by race. Hypotheses are tested related to each specific aim based on a multi-stage, multi-level conceptual model reflecting the influence of race/ethnicity on the inputs children bring to the process of mental health services and special education participation, as well as its influence on the process itself. Extensive data gathered annually since 1990 on non-intervention participants in the Fast Track project are analyzed. These data are uniquely suited to the proposed project because they include detailed information from school and medical records as well as parental interviews providing information on mental health functioning, symptomatology, and diagnosis. [unreadable] [unreadable] Analyses to be performed include Kaplan-Meier survival functions to describe the timing of entry and exit from special education and mental health services; multivariate Cox proportional hazards models and multinomial logit models to examine the effects of child and family characteristics on timing of services and restrictiveness of educational placement, respectively; and multi-level hazard models quantifying the influence of school and community characteristics on disparities taking individual-level factors into account. [unreadable] [unreadable]